Skip to content

rust.sqlx.missing_pool_timeout

Stability Medium

Detects SQLx pool configuration without connection timeout.

Why It Matters

Without pool timeouts:

Hanging requests — Waiting for connections that never come
Connection starvation — Pool exhaustion blocks all queries
Cascade failures — Database issues spread to your service
No feedback — Can’t tell why things are slow

Pool timeouts provide fail-fast behavior and observability.

Example

// ❌ Before
let pool = PgPoolOptions::new()
    .max_connections(5)
    .connect("postgres://...")
    .await?;

// ✅ After
use std::time::Duration;

let pool = PgPoolOptions::new()
    .max_connections(5)
    .acquire_timeout(Duration::from_secs(3))
    .connect("postgres://...")
    .await?;

What Unfault Detects

PgPoolOptions without acquire_timeout
Missing connection timeout
Pool without max connections set

Auto-Fix

Unfault adds acquire_timeout(Duration::from_secs(3)) to pool options.

Complete Pool Configuration

use sqlx::postgres::PgPoolOptions;
use std::time::Duration;

let pool = PgPoolOptions::new()
    // Connection limits
    .max_connections(10)
    .min_connections(2)

    // Timeouts
    .acquire_timeout(Duration::from_secs(3))
    .idle_timeout(Duration::from_secs(600))
    .max_lifetime(Duration::from_secs(1800))

    // Connection testing
    .test_before_acquire(true)

    .connect(&database_url)
    .await?;

Timeout Guidelines

Setting	Purpose	Suggested Value
`acquire_timeout`	Wait for connection from pool	3-5 seconds
`idle_timeout`	Close unused connections	10 minutes
`max_lifetime`	Recycle connections	30 minutes